Windows Vista question

[digitalcraft]

oh yeah ? for 70,000 desktops ? No didn't think so.
Until the linux community delivers reliable centrlaised image management then it has a ways to go in industry. It is fine for servers and where businesses are using it. But has a way to go for the big systems .... but will get there

If a company is running 70,000 desktops it will have a few central servers too. Linux will save a company like that tons of money. Oregon State University has quite a few linux labs, and they integrate with the other labs. If you can do it for labs, you can multiply that for a larger scale. Switches and such work fine with linux, I don't see how this is more complex

[NuclearCrap]

I see that the main issue of Vista is security and compatibility, not requirements. Of course lower-spec machines can just run the classic theme. I had Vista Beta installed on my old computer a while ago (Pentium III, 512MB RAM, FX5200) and it ran as fast as XP in terms of desktop performance, but I couldn't run most of the apps.

Nowadays, we have become used to Moore's Law so know that in 18 months it will be twice as fast, twice as big and so why spend valuable development $$$$ when the problem goes away so quickly.

It was 24 months, but scientists working on CPUs said that they would probably take more than 24 months to double the number of transistors since they're about to hit a wall soon. They're at 45nm and 22nm is as far as they can go with silicon, unlike the graphics card industry which doubles the power like every every 6 months.......

And yeah hardware power is no problem seeing how fast the industry is moving as you said, unless when it comes to parents who upgrade like once every decade...........Lol.

[TheOne]

i ran vista on a p2 500mhz with 384mb ram, and some old radeon 7500, it ran, but it ran like s***, but i was happy it ran.....(this was a beta copy btw)

vista has many other things that run on the background + its fancy graphics and indexing service, if you turn off some of that stuff, vista should be able to run on slower older pc's.

also 1 of the things vista is takin ram on, is it checks for DRM every 10 secs or so.(DRM as in copyright stuff)

[Matra et Alpine]

If a company is running 70,000 desktops it will have a few central servers too. Linux will save a company like that tons of money. Oregon State University has quite a few linux labs, and they integrate with the other labs. If you can do it for labs, you can multiply that for a larger scale. Switches and such work fine with linux, I don't see how this is more complex

It was ONLY the desktops as that's where the voluem of issues expands exponentially and costs of doing it inefficiently hurt worst.
You can't compare a few "labs" with an integrated business model for a large mutinational. It's not about "making them talk" but about ensuring they are consistent and reliable for business-critical use by all employees.
As I'd said, Linux isn't YET there but some of the stuff in SUSE is making it better and will get their in the end.

[Matra et Alpine]

It was 24 months

Being pedantic -- ME ? Never - Intel said the count was 18 months and the performance 20 months. Some areas were a year.
I have tended to see the technology I was in move faster so I am guilty of stating it as 18 ( mid point ). Correct tho' Moore said 2 years.

but scientists working on CPUs said that they would probably take more than 24 months to double the number of transistors since they're about to hit a wall soon. They're at 45nm and 22nm is as far as they can go with silicon, unlike the graphics card industry which doubles the power like every every 6 months.......

Yeah graphics designers are still on the development of the algorithms and h/w implementations. They will get to the wall quicker than it took alu and bus designers.
re the hardware tho' even it has innovations and the new HP crossbar construction has lots of promise to keep things going. "Rocks Law" predicted a slow done of Moores as the cost of fab plants is doubling every 4 years, so whilst we think we can get to 4nm it is liekly to take a long time due to those cost inhibitors.